---
title: "Analysis report for criminal incidents in Victoria"
output: 
  flexdashboard::flex_dashboard:
        storyboard: true
        vertical_layout: scroll
        orientation: rows
        source_code: embed
---

```{r setup, include=FALSE}
library(flexdashboard)
library(tidyverse)
library(readxl)
library(haven)
library(ggplot2)
library(kableExtra)
library(ggResidpanel)
library(bookdown)
library(plotly)
library(here)
library(dplyr)
library(naniar)
library(tidytext)
library(knitr)
```

Introduction {data-icon="fa-address-book"}
===================================== 

```{r , fig.align ="center"}
include_graphics("police.jpeg") 
```

Row
----
### **Data Description**

The [data](https://www.crimestatistics.vic.gov.au/crime-statistics/latest-victorian-crime-data/download-data-0) used for the analysis records the criminal incidents in Victoria from 2010 December to 2020 December.  It is collected from the [Crime Statistics Agency](https://www.crimestatistics.vic.gov.au/about-us/contact-us) of Victoria Government. The copyright is under the [CC-BY 4.0 license](https://creativecommons.org/licenses/by/4.0/), which allows the team to apply the data analysis with minimum restrictions on the use of the data.

The **purpose of this analysis report** is to assist the policy maker to adjust or re-assign the police resources, possibly providing some insights on the future development on the social security.

In the data analysis, it contains three main parts:

- 1 To obtain an overview of the criminal incidents among Victoria's police regions. Also, observing the changes of incidents recorded over time to seek some trend or pattern in different local goverment areas.

  - [ ] Which police region has the greatest number of incidents recorded?
  - [ ] How did the trend of incidents recorded in those police regions varied across the years?
  - [ ] What is the trend of incidents recorded in North West Metro's local government areas changes over time?

- 2 To understand the criminal activity better, it will not only show the record changes over time but also the changes of offence sub-groups and sub-divisions. This will provide a precise landscape of criminal activities in Victoria.

  - [ ] For each LGA , which offence_subdivision, recorded maximum incidents
  - [ ] For each LGA,Which offence_subgroup was most recorded in the previous found offence_subdivision's?
  - [ ] What was the trend of each offence_subgroup found over the years?

- 3 Building on the previous analysis, this part will explore the dissect the criminal activities among different suburbs or towns in Victoria.

  - [ ] For each Suburb/Town, which offence_subdivision, recorded maximum incidents
  - [ ] For each Suburb/Town,Which offence_subgroup was most recorded in the previous found offence_subdivision's?

Row
-----

***

**Presentation TEAM**

|Name	          |Email Id	                  |Student Id|
|---------------|:-------------------------:|----------|
|Karan Grag     |kgar0017@student.monash.edu|  |
|Jiaying Zhang  |jzha0342@student.monash.edu|30930685  |
|Ziyao Wang     |zwan0008@student.monash.edu|27925315  |

 
***



Section 1 {data-icon="fa-github"}
===================================== 
```{r cleandata, include=FALSE}
dat <- read_excel("data/Data_Tables_LGA_Criminal_Incidents_Year_Ending_December_2020.xlsx", sheet=2)
dat <- dat %>% 
  select(c(Year, `Police Region`, `Local Government Area`, `Incidents Recorded`)) 
```
```{r readdata, include=FALSE}
dat1 <- read.csv("data/LGA_Criminal_data2020.csv")
```

Row {data-width=1000}
------
### Table 1 - To observe the total number of incidents recorded for each police region
    
```{r table1, message=FALSE}
dat_tot <- dat1 %>% 
  filter(`Local.Government.Area` == "Total") %>% 
  group_by(`Police.Region`) %>% 
  summarise(Total_Incidents = sum(`Incidents.Recorded`)) %>% 
  arrange(-Total_Incidents)
table1 <- dat_tot %>% 
  knitr::kable(caption = "Number of Incidents Recorded in different Police Region", align = 'c') %>% 
  kable_styling(bootstrap_options = c("striped", "hover","basic"))
table1
```
---

- **North West Metro** police region has recorded maximum incidents during the period.

Row {.tabset data-width=1000}
-------------------------------------

### Chart 2 - Incidents reocrded by different police region over time.

```{r total,fig.cap = "Total number of incidents recorded in each police region", fig.height=8, fig.align='center',fig.width=10}
dat_var <- dat1 %>% 
  filter(`Local.Government.Area` == "Total") %>% 
  group_by(Year, `Police.Region`) %>% 
  summarise(Total_Incidents = sum(`Incidents.Recorded`))
figure1 <- ggplot(dat_var, aes(x= Year,
                               y = Total_Incidents,   
                               color = `Police.Region`))+
  geom_line() +
  geom_point() +
  scale_x_continuous(breaks=seq(2011,2020,2)) +
  theme_bw()+
  facet_wrap(~ `Police.Region`, scales = "free",ncol=2)+
  scale_fill_brewer(palette = "Dark2")
ggplotly(figure1)
```

---
- In the four main police regions of Victoria, it illustrates an evident fact that the number of incidents recorded has reached the *peak around early 2016*, then has a sudden pungle afterward.

### Chart 3 - Crimial activity in North West Metro area.
    
```{r north, fig.cap = "Crimial activity trend in different suburbs from North West Metro", fig.align='center', fig.height=11,fig.width=10}
dat_North <- dat1 %>% 
  filter(Police.Region == "1 North West Metro") %>% 
  select(-X)
trend <- dat_North %>% 
  group_by(Local.Government.Area, Year) %>% 
  summarise(Total_Incidents = sum(`Incidents.Recorded`)) %>% 
  arrange(-Total_Incidents) %>% 
  filter(Local.Government.Area %in% c("Melbourne", 
                                      "Hume", 
                                      "Brimbank", 
                                      "Wyndham", 
                                      "Whittlesea", 
                                      "Moreland",
                                      "Banyule",
                                      "Darebin",
                                      "Hobsons Bay",
                                      "Maribyrnong",
                                      "Melton",
                                      "Moonee Valley",
                                      "Nillumbik",
                                      "Yarra"))
figure2 <- ggplot(trend, aes(x= Year, 
                             y = Total_Incidents,
                             color = Local.Government.Area))+
  geom_line()+
  geom_point() +
  scale_x_continuous(breaks=seq(2011,2020,2)) +
  facet_wrap(~ `Local.Government.Area`, scales = "free",ncol=2)+
  scale_fill_brewer(palette = "Dark2")+
  theme_bw()
ggplotly(figure2)
```


Section 2 {data-icon="fa-github-alt"}
===================================== 
---
### The definition of some concepts to be used

```{r , fig.align ="center"}
include_graphics("type.png") 
```


```{r read-data, include=FALSE}
criminaldata <- read_excel("data/Data_Tables_LGA_Criminal_Incidents_Year_Ending_December_2020.xlsx", sheet = 4)
```
```{r datacleaning, include=FALSE}
criminaluse <- criminaldata %>% 
  select(Year,
         `Local Government Area`,
         `Offence Subdivision`,
         `Incidents Recorded`)
```

### Which offence_subdivision, recorded maximum incidents in each LGA
Row {.tabset data-width=1000}
-------------------------------------
### Chart1 
```{r vis1, fig.width=9,fig.height=12}
  criminalfinal<- criminaluse %>% 
    mutate(lgth = str_length(`Offence Subdivision`)) %>% 
  mutate(`Offence Subdivision` = substr(`Offence Subdivision`,start = 5,stop = lgth)) %>% 
  mutate(`Offence Subdivision` = str_replace(`Offence Subdivision`,"r crimes against the person","Other crimes against the person")) %>% 
  group_by(`Local Government Area`, `Offence Subdivision` ) %>% 
    summarise(incidents = sum(`Incidents Recorded`)) 
data<- criminalfinal %>% arrange(desc(incidents)) %>%
slice(1) %>% 
ggplot(aes(y=reorder(`Local Government Area`,incidents),
       x=incidents,
       fill = `Offence Subdivision`)) +
  geom_col()+
  geom_text(aes(label = incidents)) +
  xlab("No. of Incidents") +
  ylab("Local Government Area") +
  theme(axis.text.x = element_blank()) 
 ggplotly(data)
```


### Chart2 

```{r vis2, fig.width=10}
  criminalfinal2<- criminaluse %>% 
   mutate(lgth = str_length(`Offence Subdivision`)) %>% 
  mutate(`Offence Subdivision` = substr(`Offence Subdivision`,start = 5,stop = lgth)) %>% 
  mutate(`Offence Subdivision` = str_replace(`Offence Subdivision`,"r crimes against the person","Other crimes against the person")) %>% 
  group_by(Year,
           `Offence Subdivision` ) %>% 
    summarise(incidents = sum(`Incidents Recorded`)) %>% 
ggplot(aes(x=as.numeric(Year),
       y=incidents,
       color = `Offence Subdivision` )) +
  geom_line()+
  geom_point() +
  scale_x_continuous() +
  ylab("No. of Incidents") +
  scale_x_continuous(breaks=seq(2011,2020,2)) +
  xlab("Year") 
 ggplotly(criminalfinal2) %>% 
   hide_legend()
```

Column {data-height=800, data-width = 1000}
--------------------------------

- Theft was the **most common** crime in each region.

- **Melbourne** in particular recording round 90,900 burglaries in the past five years,far more than any other region.

- Theft cases reaching a **peak** (about 156,000) in **2016**, but the number has declined in the past two years.

- Burlary/Break and Enter (about 38,000) and Assault and Related Offence (about 37,000) rank second and third, respectively, but far more lower than theft case.


Column {data-height=800, data-width = 1000}
--------------------------------

### Chart3 Which offence_subgroup was most recorded in the previous found offence_subdivision’s?

```{r datacleaning2, include=FALSE}
criminaluse2 <- criminaldata %>% 
  select(Year,
         `Offence Subgroup`,
         `Offence Subdivision`,
         `Incidents Recorded`) %>% 
  filter(`Offence Subdivision` == "B40 Theft")
```
```{r vis3, fig.width=9,fig.cap="Incidents in each Offence_subgroup of most recorded Offence_subdivision"}
    criminalfinal3<- criminaluse2 %>%
  group_by(Year, `Offence Subgroup`) %>%
    summarise(incidents = sum(`Incidents Recorded`)) %>%
ggplot(aes(Year,
       incidents,
       fill = `Offence Subgroup` )) +
  geom_col(position="dodge")+
   ggtitle("Incidents in each Offence_subgroup of most recorded Offence_subdivision")
 ggplotly(criminalfinal3)
```
- Among thefts, **steal from a motor vehicle** has the largest number during the last five years. In 2016, about 63,000 steal from a motor vehicle occurred, but the number of thefts has decreased in the past two years.

- **Other Theft** is second (around40000), and other offence_subgroup make up only a small fraction.


Row {data-height=600, data-width = 1000}
--------------------------------
### Chart4 What was the trend of each offence_subgroup over the years?

```{r datacleaning3, include=FALSE}
criminaluse3 <- criminaldata %>% 
  select(Year,
         `Offence Subgroup`,
         `Incidents Recorded`)
```
```{r vis4, fig.width=10}
  criminalfinal4<- criminaluse3 %>% 
  mutate(lgth = str_length(`Offence Subgroup`)) %>% 
  mutate(`Offence Subgroup` = substr(`Offence Subgroup`,start = 5,stop = lgth)) %>% 
  mutate(`Offence Subgroup` = str_replace(`Offence Subgroup`,"r crimes against the person","Other crimes against the person")) %>% 
  group_by(Year, `Offence Subgroup`) %>% 
    summarise(incidents = sum(`Incidents Recorded`)) %>% 
  arrange(`Offence Subgroup`,desc(incidents)) %>% 
    ggplot(aes(Year,
       incidents,
       color = `Offence Subgroup`)) +
  geom_line()+
  scale_x_continuous(breaks=seq(2011,2020,2))+
  geom_point() +
  ylab("No. of Imcidents")
 ggplotly(criminalfinal4) %>% 
   hide_legend()
```
Row {data-height=800, data-width = 1000}
--------------------------------
- Most offence_subgroup have seen a **decline** in crime over the **last year**, and almost all have seen a small spike in **2016**.

- **Steal from a motor vehicle** in the category of theft has been in the first place in the past five years, with about 50,000 cases occurring every year, far exceeding other offence_subgroups.

- **Other Theft** is in second place (about 39,000 per year) and **Criminal Damage** is third (about 35,000 per year). However, the gap between these two types of crimes and steal from a notor vehicle has gradually **widened** since 2013.

- Since 2019,**public health and safety** cases have **risen** sharply, with almost no cases in 2019 but reached about 37,000 in **2020**.



Section 3 {data-icon="fa-gitlab"}
===================================== 
```{r read-file,include=FALSE}
data3 <- read_excel(here::here("data/Data_Tables_LGA_Criminal_Incidents_Year_Ending_December_2020.xlsx"),sheet = 4)
```

```{r clean,include=FALSE}
miss_var_summary(data3)
data3 <- data3 %>% 
  rename(Month = `Year ending`,
         LGA = `Local Government Area`,
         Suburb = `Suburb/Town Name`,
         Offence_Division = `Offence Division`,
         Offence_Subdivision = `Offence Subdivision`,
         Offence_Subgroup = `Offence Subgroup`,
         Incidents_Recorded = `Incidents Recorded`)
```

Row {data-width=650}
---

### Table

```{r Q1-table,echo=FALSE}
data3 %>% 
  group_by(Suburb) %>%
  summarise(Total_Incidents = sum(Incidents_Recorded)) %>% 
  slice_max(Total_Incidents,n = 10) %>% 
  kable(caption = "Suburbs with maximum incidents over the years") %>% 
  kable_styling(bootstrap_options = c("striped","hover","basic"))
```
* **Melbourne** had maximum no of incidents with more than 1.5 lakh over the years 
* Other suburbs were 60K incidents in total. 

Row {data-width=1000}
---
### Chart 1

```{r Q1,echo=FALSE,fig.width=8,fig.height=15,fig.cap="Top 10 Suburb with most incidents recorded"}
data3 %>% 
  group_by(Year,Suburb) %>%
  summarise(Total_Incidents = sum(Incidents_Recorded)) %>% 
  arrange(Year,desc(Total_Incidents)) %>% 
  slice_max(Total_Incidents,n = 10) %>% 
  mutate(Suburb1 = reorder_within(Suburb,Total_Incidents,Year)) %>% 
  ggplot(aes(x=Total_Incidents ,
             y=Suburb1,
             fill = Suburb)) +
  geom_col() +
  geom_text(aes(label = Total_Incidents)) +
  scale_y_reordered() +
  ylab("Suburb") +
  xlab("No. of Incidents") +
  ggtitle("Top 10 Suburb with most incidents recorded in each Years") +
  facet_wrap(~Year,ncol = 1, scales = "free")
```

Row {data-height=200, data-width = 1000}
--------------------------------
*  **Melbourne** not only had overall maximum incidents, but also had the maxiimum crime in each year.
*  It had least no of crimes in year **2015**.
*  Suprisiingly, **Melbourne** saw a massive fall of 10K in crime no. from 2019-2020, mqybe due to lockdown restrictions imposed  during COVID-19.


Row {data-width=1000}
--------------------------------
### Chart 2

```{r Q2,echo=FALSE,fig.width=12,fig.height=15,fig.cap="Top 10 Offences recorded"}
data3 %>%
  mutate(lgth = str_length(Offence_Subdivision)) %>% 
  mutate(Offence_Subdivision = substr(Offence_Subdivision,start = 5,stop = lgth)) %>% 
  mutate(Offence_Subdivision = str_replace(Offence_Subdivision,"r crimes against the person","Other crimes against the person")) %>% 
  group_by(Year,Offence_Subdivision) %>% 
  summarise(Tot_incidents = sum(Incidents_Recorded)) %>% 
  arrange(Year,desc(Tot_incidents)) %>%
  slice_max(Tot_incidents,n = 10) %>% 
  mutate(Offence1 = reorder_within(Offence_Subdivision,Tot_incidents,Year)) %>% 
    ggplot(aes(y= Offence1,
         x= Tot_incidents,
         fill = Offence_Subdivision)) +
  geom_col() +
  geom_text(aes(label = Tot_incidents)) +
  scale_y_reordered() +
  ylab("Type of Offece") +
  xlab("No of Incidents") +
  facet_wrap(~Year,ncol = 1,scales = "free")
```
Row {data-height=800, data-width = 1000}
--------------------------------
* **Theft** was the most occurred crime across all the years.
* **Burglary/Break and enter** type of offence decreased over the years.

Row {data-width=1000}
--------------------------------

### Chart 3
```{r Q3,echo=FALSE,fig.width=10,fig.height=4,fig.cap="Top 10 Offfences Suburrb wise"}
Suburb_imp <- data3 %>% 
  group_by(Year,Suburb) %>%
  summarise(Total_Incidents = sum(Incidents_Recorded)) %>% 
  arrange(Year,desc(Total_Incidents)) %>% 
  slice_max(Total_Incidents,n = 10) 
Q3graph <- data3 %>% 
  mutate(lgth = str_length(Offence_Subdivision)) %>% 
  mutate(Offence_Subdivision = substr(Offence_Subdivision,start = 5,stop = lgth)) %>% 
  mutate(Offence_Subdivision = str_replace(Offence_Subdivision,"r crimes against the person","Other crimes against the person")) %>% 
  filter(Suburb %in% unique(Suburb_imp$Suburb)) %>% 
  group_by(Suburb,Offence_Subdivision) %>% 
  summarise(Tot_incidents = sum(Incidents_Recorded)) %>% 
  slice_max(Tot_incidents,n = 2) %>% 
  arrange(-Tot_incidents) %>% 
  #mutate(offence1 = reorder_within(Offence_Subdivision,Tot_incidents,Suburb)) %>% 
  ggplot(aes(y= Suburb,
             x =Tot_incidents,
             fill = Offence_Subdivision)) +
  geom_col() 
  ggplotly(Q3graph)
```

* **Theft** is most occurred crime across the affected Suburbs too

Row {data-width=1000}
--------------------------------
### Chart 4
```{r Q4,echo=FALSE,fig.width=10,fig.height=10,Fg.cap="Trend of Top 2 Offences in each year w.r.t Suburb" }
Q4graoh <- data3 %>% 
  mutate(lgth = str_length(Offence_Subdivision)) %>% 
  mutate(Offence_Subdivision = substr(Offence_Subdivision,start = 5,stop = lgth)) %>% 
  mutate(Offence_Subdivision = str_replace(Offence_Subdivision,"r crimes against the person","Other crimes against the person")) %>% 
  filter(Suburb %in% unique(Suburb_imp$Suburb)) %>% 
  group_by(Year,Suburb,Offence_Subdivision) %>% 
  summarise(Tot_incidents = sum(Incidents_Recorded)) %>% 
  slice_max(Tot_incidents,n = 2) %>% 
  ggplot(aes(x= as.numeric(Year),
             y =Tot_incidents,
             color = Offence_Subdivision)) +
  geom_line() +
  geom_point() +
  scale_x_continuous(breaks=seq(2011,2020,2)) +
  xlab("Year") +
  ylab("Total_incidents") +
  facet_wrap(~Suburb)
ggplotly(Q4graoh)
```
Row {data-width = 1000}
--------------------------------
*  Crime rate of the 2 most occured offences im the affected Suburbs were pretty much steafy along the years.
*  Im **Melbourne**.
   *  **Disorderly and offensive conduct** crime went off radar after 2014.
   *  **Breachers of order** were only seen from years 2015-2019.
   *  **Miscellaneous offence** came into action in year 2020.

Conclusion & References {data-icon="fa-table"}
=====================================     


Column {data-height=800, data-width = 1000}
--------------------------------
### Conclusion:

-- In part two, the report show that 

- In every LGA, **theft** occurs most frequently in offence_subdivisions and the incidence of thefts is far greater than other types of offence_subdivision.

- In the theft offence_subdivision, the vast majority are **steal from a motor vehicle** and **other theft**. 

- Because the proportion of theft cases is too large, steal from a motor vehicle and other Theft also the top two in all offence_subgroup. 

- Overall, most of offence_subgroups's number of incidents were up in **2016**, but have declined in the last two years.




***

**References**
[1] Crime Statistics Agency Victoria. (2021). Victim Reports. Retrieved from https://www.crimestatistics.vic.gov.au/crime-statistics/latest-victorian-crime-data/victim-reports-2

[2]  C. Sievert. Interactive Web-Based Data Visualization with R, plotly, and shiny. Chapman and Hall/CRC Florida, 2020.

[3] Hadley Wickham and Evan Miller (2021). haven: Import and Export 'SPSS', 'Stata' and 'SAS' Files. R package version 2.4.1.
  https://CRAN.R-project.org/package=haven

[4] Hadley Wickham and Jennifer Bryan (2019). readxl: Read Excel Files. R package version 1.3.1. https://CRAN.R-project.org/package=readxl

[5] Hao Zhu (2021). kableExtra: Construct Complex Table with 'kable' and Pipe Syntax. R package version 1.3.4.
  https://CRAN.R-project.org/package=kableExtra

[6] H. Wickham. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York, 2016.

[7] Katherine Goode and Kathleen Rey (2019). ggResidpanel: Panels and Interactive Versions of Diagnostic Plots using 'ggplot2'. R package
  version 0.3.0. https://CRAN.R-project.org/package=ggResidpanel

[8] Kirill Müller (2020). here: A Simpler Way to Find Your Files. R package version 1.0.1. https://CRAN.R-project.org/package=here

[9] R Core Team (2021). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL
  https://www.R-project.org/.

[10] Richard Iannone, JJ Allaire and Barbara Borges (2020). flexdashboard: R Markdown Format for Flexible Dashboards. R package version 0.5.2.
  https://CRAN.R-project.org/package=flexdashboard

[11] Yihui Xie (2021). bookdown: Authoring Books and Technical Documents with R Markdown. R package version 0.22.

[12] Yihui Xie (2016). bookdown: Authoring Books and Technical Documents with R Markdown. Chapman and Hall/CRC. ISBN 978-1138700109


***

Thank You!